S100 and S100β: biomarkers of cerebral damage in cardiac surgery with or without the use of cardiopulmonary bypass.

OBJECTIVE: The present study is to describe the clinical impact of S100 and S100β for the evaluation of cerebral damage in cardiac surgery with or without the use of cardiopulmonary bypass (CPB).
METHODS: Quantitative results of S100 and S100β reported in the literature of the year range 1990-2014 were collected, screened and analyzed.
RESULTS: Cerebrospinal fluid and serum S100 levels showed a same trend reaching a peak at the end of CPB. The cerebrospinal fluid/serum S100 ratio decreased during CPB, reached a nadir at 6 h after CPB and then increased and kept high untill 24 h after CPB. Serum S100 at the end of CPB was much higher in infant than in adults, and in on-pump than in off-pump coronary artery bypass patients. ∆S100 increased with age and CPB time but lack of statistical significances. Patients receiving an aorta replacement had a much higher ∆S100 than those receiving a congenital heart defect repair. Serum S100β reached a peak at the end of CPB, whereas cerebrospinal fluid S100 continued to increase and reached a peak at 6 h after CPB. The cerebrospinal fluid/serum S100β ratio decreased during CPB, increased at the end of CPB, peaked 1 h after CPB, and then decreased abruptly. The increase of serum S100β at the end of CPB was associated with type of operation, younger age, lower core temperature and cerebral damages. ∆S100β displayed a decreasing trend with age, type of operation, shortening of CPB duration, increasing core temperature, lessening severity of cerebral damage and the application of intervenes. Linear correlation analysis revealed that serum S100β concentration at the end of CPB correlated closely with CPB duration.
CONCLUSION: S100 and S100β in cerebrospinal fluid can be more accurate than in the serum for the evaluations of cerebral damage in cardiac surgery. However, cerebrospinal fluid biopsies are limited. But serum S100β and ∆S100β seem to be more sensitive than serum S100 and ∆S100. The cerebral damage in cardiac surgery might be associated with younger age, lower core temperature and longer CPB duration during the operation. Effective intervenes with modified CPB circuit filters or oxygenators and supplemented anesthetic agents or priming components may alleviate the cerebral damage.

INTRODUCTION

S100 protein family members with a molecular mass of 10-12 kDa are acidic proteins characterized by their calcium-dependent biological effects[. It is expressed in different tissues, but shows brain tissue specific, and therefore implicated in cerebral damage. They may form into homodimers, heterodimers and even oligomers based on a calcium-dependent conformational change[. Most S100 proteins have a low binding affinity for calcium, which increase dramatically to control a cellular activity in the presence of a target[. This protein family represents the largest subgroup within the superfamily of EF-hand Ca2+ binding proteins. Ca2+ binding to the first EF-hand (helix I, loop and helix II) is weaker than binding to the second EF-hand (helices III and IV)[. S100β, a 10.7 kDa protein, is a member of S100 protein family. It is highly expressed in astrocytes and is one of the most abundant soluble proteins in human brain, constituting 0.5% of them. S100β functions as both an intracellular Ca2+ receptor and an extracellular neuropeptide by way of the receptor for advanced glycation end-products, a main transducer of extracellular functions of this protein[. S100β is displayed as a homodimer with a high binding affinity under all biological circumstances while the monomers are absent[.

Blood-brain barrier dysfunction secondary to cerebral damages may expedite the release of these cerebral specific proteins from the astroglial or Schwann cells into cerebrospinal fluid (CSF) and blood circulation[. During cardiac operations, neurological disorders may occur and are believed to be the results of thromboembolism (embolism is not always caused by a thrombus, but can be air embolism, calcium embolism or detachment of atheromatous plaques from the aorta at the time of cannulation or decannulation) and systemic inflammatory reactions[. S100 and S100β have been reliable serum markers of cerebral damage due to breakdown of the blood-brain barrier caused by head trauma, anoxia, ischemia, neoplasm and cardiac surgery[. Both hypo- and hypertension may also cause cerebral damage by impairment of cerebral autoregulation[. S100 and S100β proteins leak from structurally damaged neurocytes into CSF and then across the blood-brain barrier. S100β protein increases 50~100-fold after cardiac operation with cardiopulmonary bypass (CPB), supporting links between CPB, microembolization and cerebral damage[ and indicating postoperative adverse neurologic outcomes[. However, debates remain with regard to the accuracy of the results during and early after the operation as well as the correlations between the expression of the proteins and the surgical conditions. In order to highlight these aspects, a comprehensive review is made based on quantitative data reported in the literature.

METHODS

Literature Retrieval

A literature search for English articles published from 1990 to 2012 concerning S100 and S100β in relation to cardiovascular surgery in PubMed, Highwire Press and Google search engine yielded totally 69 publications[. The search terms included "S100", "S100β(B)", "cardiopulmonary bypass" "off-pump coronary artery bypass", "circulatory arrest, induced", "profound hypothermic circulatory arrest", "cardiac surgery", "congenital heart defects", "heart valves", "coronary artery bypass grafting", "aortic surgery" and "cardiac surgical procedures". Quantitative data of S100 and S100β measured in the unit of µg/L were screened, collected and analyzed. Articles or patient cohorts reported in articles with no quantitative data were excluded from this study.

Sampling

Sampling times were before operation (baseline) (T0), during CPB (T1), at the end of CPB (T2), 1, 4, 6, 12, 24, 48, and 72 h after operation (T3-9).

Indicators

The indicators of evaluating the cerebral damage included dynamics of CSF and serum S100(β), ∆S100(β), i.e., the difference between peak and baseline S100(β)[ and CSF/serum S100(β) ratios.

Subgroups

1) Age: There were 4 age subgroups: neonate, infant, child and adult;

2) Operation: The operations were classified as aorta, valve, congenital heart defect, coronary artery bypass grafting (CABG) and off-pump coronary artery bypass (OPCAB);

3) CPB duration: There were 2 subgroups based on whether the CPB duration was >100 minutes;

4) Core temperature: There were 3 subgroups according to core temperatures during CPB: deep hypothermia, mild and moderate hypothermia and normothermia;

5) Cerebral damage: The patients with cerebral damage were divided into either functional (confusion, agitation, disorientation, or epileptic seizures) or organic (stroke, stupor, or coma) subgroups. Those without cerebral damage were defined as control; and,

6) Intervene: Patients with utilizations of modified CPB circuit and oxygenators[, cell saving reservoir[, anesthetic agents and priming components (propofol[, isoflurane[, hydroxyethyl[ and starch[) during the operation aiming at lessening the cerebral damage were defined as the Intervene Subgroup. Those without intervenes were defined as control.

Statistical analysis

Data were expressed as mean±standard deviation. Comparisons between groups were conducted with unpaired t -test, and linear correlations were assessed between independent and dependent variables. P <0.05 was considered statistically significant.

RESULTS

Patient information

The 69 articles reported the quantitative results of S100(β) of 4439 patients: 20 (29.0%) on serum S100[, 45 (65.2%) on serum S100β[, 2 (2.9%) on serum and CSF S100[, 1 (1.4%) on serum and CSF S100β[ and 1 (1.4%) on CSF S100β[. The 2 articles reporting CSF S100 comprised 22 patients with 15 males and 6 females with a median age of 63 years. All received a thoracic aorta operation with postoperative spinal cord injury in 2 (9.1%) patients; and the 2 articles reporting CSF S100β included 49 patients with 28 males and 23 females (gender of 8 patients was unidentified) with a median age of 64 years. All received a thoracic aorta operation with postoperative spinal cord injury in 10 (20.4%) patients. The demographics of the patients with serum S100(β) detections were listed in Table 1.

Table 1

Demographics of patients with serum S100 and serum S100ß detections.

Variable	S100	S100β
Report number	20	45
Patient number	1741	2682
Gender (male/female)	1217/352	829/451
Age	51.8±26.1 (range, 3 days-77 years; median, 68 years)	47.7±27.8 (range, 8.6 days-81 years; median, 62 years)
Operative conditions
CPB (min)	98.1±36.5 (range, 48-217; median, 90.3)	116.5±55.0 (range, 49-308; median, 103.2)
Crossclamp time (min)	63.8± 31.5 (range, 29-175; median, 56)	64.9±29.9 (range, 28-164; median, 60)
Hypothemic circulatory arrest time (min)	43.1±24.2 (range, 26-60.2; median, 43.1)	31.6±9.1 (range, 20-45; median, 32)
Core temperature (T)	30.1±4.4 (range, 18-34.5; median, 31.5)	29.46±6.5 (range, 10.5-37; median, 32)
Age group, n (%)
Neonate	25 (1.4)	173 (6.5)
Infant	17 (1.0)	69 (2.6)
Child	21 (1.2)	18 (0.7)
Adult	1678 (96.4)	2422 (90.3)
Core temperature, n (%)
Deep hypothermia	44 (2.6)	278 (10.4)
Mild-moderate hypothermia	1576 (93.9)	2250 (83.9)
Normothermia	58 (3.5)	154 (5.7)
Operation, n (%)
Aorta replacement	31 (1.8)	192 (7.1)*
Valve replacement	14 (0.8)	156 (5.8)
Congenital heart defect repair	64 (3.7)	270 (10.0)
CABG	1335 (76.7)	1941 (72.2)
OPCAB	229 (13.2)	129 (4.8)
Not given	68 (3.9)
Cerebral damage, n (%)	23 (1.3)	121 (4.5)
Organic cerebral damage	23 (100)	65 (53.7)
Stroke	3 (13.0)	58 (89.2)
Transient ischemic attack	1 (4.3)	0 (0)
Spinal cord injury	3 (13.0)	2 (3.1)
Subclinical cerebral damage	16 (69.6)	5 (7.7)
Functional cerebral damage	0 (0)	56 (46.3)
Intervene (with modified filter, oxygenator or anesthetic agents)	259 (14.9)	330 (12.3)
Assay, n (%)
Immunoradiometry	985 (56.6)	891 (33.2)
Enzyme linked immunosorbent assay	163 (9.4)	235 (8.8)
Immunoluminometry	161 (9.2)	668 (24.9)
Immunofluorometry		500 (18.6)
Luminometry		128 (4.8)
Immunoassay		72 (2.7)
Electrochemoluminescence immunoassay		21 (0.8)
Not given	432 (24.8)	167 (6.2)

at least 5 patients had concurrent procedures. CABG=coronary artery bypass grafting; CPB=Cardiopulmonary bypass; OPCAB=off-pump coronary artery bypass

Assays

Immunoradiometry, immunoluminometry and immunofluorometry were the 3 main assays used for the detection of the biomarkers (Table 1).

Biomarkers

CSF and serum S100 levels showed a same trend during the early observational stage before T5, increased at T1, reaching a peak at T2 and then gradually decreased. After T5, CSF S100-serum S100 separation phenomenon was seen. The CSF/serum S100 ratio decreased from T1, reached a nadir at T5 and then increased and kept high till T7 (Figure 1).

Fig. 1

Dynamics of CSF S100, serum S100 and CSF/serum S100 ratio. CSF=Cerebrospinal fluid

Serum S100 at T3 was much higher in infant than in adults (2.4±1.2 µg/L vs . 0.9±1.0 µg/L, P =0.034) and in CABG patients than in OPCAB patients (2.8±2.4 µg/L vs . 0.8±0.6 µg/L, P =0.010). Patients with a CPB time >100 min had a higher serum S100 level at T2 than those with a CPB time <100 min, but lack of a statistical significance, however, significant reductions were noted at T7 in comparison to T2 in both subgroups (CPB >100 min: 3.3±2.3 µg/L vs . 0.6±0.6 µg/L, P =0.005; CPB duration <100 min: 2.1±2.3 µg/L vs . 0.3±0.2 µg/L, P =0.016). Deep hypothermia circulatory arrest was associated with much higher serum S100 at T2 than mild-moderate hypothermia and normothermia patients, and mild-moderate hypothermia with higher serum S100 than normothermia. No difference in the serum S100 levels was noted between patients with cerebral damage in particular stroke and those without. Intervenes with CPB filter, oxygenator, or anesthetic agents led to significant decreased serum S100 at T2 and T7 (Figure 2).

Fig. 2

An inter-subgroup comparison of serum S100 at T2 and T7. CABG=Coronary artery bypass; CHD=Congenital heart defect; DHCA=Deep hypothermia circulatory arrest; FCD=Functional cerebral damage; MMH=Mild-moderate hypothermia; NM=Normothermia; OCD=Organic cerebral damage; OPCAB=Off-pump coronary artery bypass; T2=At the end of cardiopulmonary bypass; T7=24 hours after cardiopulmonary bypass

∆S100 could be calculated in 25 series of patients in whom at least a baseline and a peak value were reported. The peaks were at T1 in 5 (20%), T2 in 16 (64%) and T3 in 4 (16%) patient cohorts, respectively (Χ =7.5, P =0.023). ∆S100 increased with age and CPB time but lack of statistical significances. Patients receiving an aorta replacement had a much higher ∆S100 than those receiving a congenital heart defect repair, in line with the increasing trend with age. No difference was found in ∆S100 between deep hypothermia and mild-moderate hypothermia patients or between the organic cerebral damage and control patients. Intervenes led to a decrease of ∆S100 in comparison to non-intervene patients but no significance was found (Figure 3).

Fig. 3

An inter-subgroup comparison of serum ∆S100 at T2 and T7. CABG=Coronary artery bypass; CHD=Congenital heart defect; DHCA= Deep hypothermia circulatory arrest; FCD=Functional cerebral damage; MMH=Mild-moderate hypothermia; NM=Normothermia; OCD=Organic cerebral damage; OPCAB=Off-pump coronary artery bypass

CSF and serum S100β levels started to increase at T1, but separation was noted since T2. Serum S100β reached a peak at T2, whereas CSF S100β continued to increase and reach a peak at T5. Both recovered to normal at T7. The CSF/serum S100β ratio decreased at T1, increased at T2, peaked at T3 and then decreased abruptly (Figure 4).

Fig. 4

Dynamics of CSF S100β, serum S100β and CSF/serum S100β ratio. CSF=Cerebrospinal fluid

Serum S100β at T2 showed a successive decrease in the operation subgroups in a sequence of aorta, valve, congenital, CABG and OPCAB operations. Patients with organic and functional cerebral damages showed higher S100β levels at T2 than those without. Infant showed a little bit higher serum S100β than adults, patients with CPB duration >100 min showed higher serum S100β than those with CPB duration <100 min, deep hypothermia and mild-moderate hypothermia were associated with higher serum S100β than normothermia, and intervene led to reduced serum S100β other than non-intervene, but no significances were found (Figure 5).

Fig. 5

An inter-subgroup comparison of serum S100β at T2 and T7. CABG=Coronary artery bypass; CHD=Congenital heart defect; DHCA=Deep hypothermia circulatory arrest; FCD=Functional cerebral damage; MMH=Mild-moderate hypothermia; NM=Normothermia; OCD=Organic cerebral damage; OPCAB=Off-pump coronary artery bypass; T2=At the end of cardiopulmonary bypass; T7=24 hours after cardiopulmonary bypass

∆S100β could be calculated in 51 series of patients. The peak values were present at T1 in 5 (9.8%), T2 in 36 (70.6%) and T3 in 10 (19.6%) patient cohorts, respectively (Χ =48.9, P =0.000): ∆S100β displayed a decreasing trend with age, surgical operations (from aorta, valve, congenital, CABG to OPCAB), shortening of CPB duration, increasing core temperature, lessening severity of cerebral damage and the application of intervenes. Significant differences were present in age, surgical operation, core temperature and cerebral damage subgroups (Figure 6).

Fig. 6

An inter-subgroup comparison of serum ∆S100β at T2 and T7. CABG=Coronary artery bypass; CHD=Congenital heart defect; DHCA=Deep hypothermia circulatory arrest; FCD=Functional cerebral damage; MMH=Mild-moderate hypothermia; NM=Normothermia; OCD=Organic cerebral damage; OPCAB=Off-pump coronary artery bypass

Linear correlation analysis did not reveal any significant correlation between serum S100 concentration at T2 and CPB, crossclamp time and core temperature (Figure 7). However, serum S100β concentration at T2 correlated closely with CPB duration (Figure 8).

Fig. 7

Linear correlation analysis between serum S100 concentration at T2 and cardiopulmonary bypass, crossclamp time and core temperature.

Fig. 8

Linear correlation analysis showed serum S100β concentration at T2 correlated closely with cardiopulmonary bypass duration.

DISCUSSION

Detectable concentrations of S100 were found 20 min after CPB[. On the operative day, CSF S100 levels increased with time for patients with spinal cord injury; whereas there was a non-specific increase of serum S100. In patients with spinal cord injury, CSF S100 was increased at 6 h after crossclamp removal[.

Serum S100 reached the peak values at the end of CPB and decreased on postoperative day (POD) 1[. At the end of the operation, S100 decreased rapidly and progressively but remained significantly higher on POD 2[. S100 peaked 20 min after the start of CPB, being significantly higher than the baseline value[ . Serum S100β increased during CPB, peaked at the late phase of CPB[, recovered to normal at 36 h after the operation[ untill POD 6[. S100β significantly increased 24 h after total circulatory arrest[.

In studies showing a correlation between neurological deficit and elevated S100β protein level after ischemic cerebral infarction, the blood level of S100β protein consistently peaked on day 2 to 3 after the clinical event[.

The release of S100β from adipose tissue with surgery would be more extensive with more complex and longer operations. These patients are at a higher risk of cerebral damage and this confounding effect may explain the correlations between early rise in S100β and neurological injury. In stroke, an elevation of S100β correlates with the amount of the damaged brain tissue. Poor neurological outcome is related to S100β levels. The peak levels of S100β occur on day 3 following the stroke[. S100β as an indicator of cerebral injury, however, is uncertain how autotransfusion of S100β from extracerebral sources is like. There is good evidence to show that autologous blood recorery through cardiotomy suckers results in significantly higher serum levels of S100β[.

Some authors have determined that shed mediastinal blood collected during surgery by cardiotomy suction contained high levels of S100β as well as chest tube blood used for autotransfusion after surgery. Therefore, early elevated serum S100β levels immediately after cardiac operations may have been contaminated by extracerebral sources of S100β[. Comparing the patients with retrograde cerebral perfusion with non-retrograde cerebral perfusion groups, the mean serum S100β levels are 0.09 and 0.09 mg/L, preoperatively, 3.8 and 4.2 mg/L 30 minutes after CPB, and 0.82 and 0.53 mg/L on POD 1[. S100β levels early after CPB are increased because of release from adipose tissue or thymus into cardiotomy suction. This masks neurally released S100β. High levels of S100β have been found in pleural drainage following thoracotomy, and in surgical wounds, mediastinal fat and skeletal muscle[. Neonates and infants had reduced S100β at 24 h after surgery than before surgery. However, this finding may reflect dilution of the protein in serum from postoperative blood, colloid and crystalloid infusions in small babies[. The increases of S100β in the early phase after cardiac surgery are not due to release of S100β from brain alone but also from tissue outside the brain[. Therefore, S100β protein is a nonspecific marker of tissue injury as glial fibrillary acidic protein might serve as a specific marker of cerebral damage after cardiac surgery[. Cerebral damage following cardiac surgery cannot be differentiated from cardiac or other tissue damage by measurement of S100β levels until the initial elevation of S100β due to non-brain tissue damage has declined, which does not occur for at least 24 h after surgery[.

It has been reported that S100 correlated significantly with age, body surface area, nasopharyngeal temperature and PaCO2 in infants and children[. However, it could be the result of dilution of the protein in serum from infusions of fluid and blood products[. Both older age and prolonged CPB duration correlated with levels of S100 protein at T0, but the correlation was weak for both variables[. Serum S100 values at the end of CPB and POD 1 significantly correlated with CPB time[. The duration of absent cerebral perfusion time (duration of circulatory arrest minus retrograde cerebral perfusion) correlated well with S100 on POD 1[.

In adults, S100 on POD 1 correlated with duration of circulatory arrest[, and peak S100β correlated with CPB time[. S100β on POD 1 correlated with duration of absent cerebral perfusion time[. S100β concentration at 5 h and 24 h correlated significantly with the duration of total circulatory arrest[ and S100β at 5 h negatively correlated with core temperature[. S100β also correlated with the total embolus count at the arterial line[, CPB time[ and intubation duration[. In roller pump group, peak S100β correlated with crossclamp time[. Ashraf et al.[ reported S100β did not correlate with duration of CPB time. Johnsson[ reported no relationship between serum S100β at 24 h after surgery and CPB duration, crossclamp time, or use of hypothermic circulatory arrest, and it did not correlate with 30-day surgical mortality.

Pulsatile flow lowers cerebral destruction than laminar flow[. S100 was nonsignificantly higher in cold than in warm CPB patients[.The S100β rise was significantly less in patients administered sevoflurane in comparison to total intravenous anesthesia[. CPB with covalent bonded heparin attached to the CPB circuit in combination with a reduced systemic heparin dose seemed to reduce the operative stroke[.

The S100 level was elevated at the end of operation but returned toward normal at 5 h. A secondary increase in S100 protein level coincided with the clinical presentation of stroke on the day after the operation[. The peak values of S100β were higher in died patients than in the survived[. Taggart et al.[ reported 21 of 43 patients had an elevated serum S100β value 4 h after the operation and none of the patients had neurological symptoms, and S100β reached a peak value on PODs 2-3 in stroke patients[. Patient with cerebral infarction showed slightly increased S100β during operation but decreased to normal concentration on POD 1. In patients with temporary left-side hemiplegia lasting 24 h after the operation, S100β protein increased and reached its peak after aortic crossclamp removal, but decreased to a normal concentration on POD 1 while still hemiplegic. In patients with a conscious disturbance lasting 24 h, S100β level was indistinguishable from the patients without neurological complications. There was a weak but significant correlation between peak concentrations of S100β protein and aortic crossclamp time in the CPB group[. The patient with the highest S100 values at the end of CPB and on POD 1 presented postoperative stroke[. Permanent cerebral damage was associated with much higher serum S100 than transient[. However, the appropriate time to measure S100β after CABG for prognostic value has not been established but is probably 5 h after surgery[.

In the hypothermic circulatory arrest group, CPB time correlated with peak S100. Peak S100β levels occurred in both the CABG and hypothermia circulatory arrest groups at the end of CPB. After 24 h, the S100β levels returned to normal in the CABG patients but were still elevated in all cases in the hypothermia circulatory arrest group. CPB patients may face major treatment-related cognitive performance decline. Persistently high levels of neuron-specific enolase might be a useful biomarker to identify patients with cognitive performance deficits at discharge; while no significant correlation between S100β levels and impaired cognitive function have been found[. High-dose propofol triggered short-term neuroprotection and long-term neurodegeneration in neuronal cultures from rat embryos[. A high dose of propofol (with plasma concentrations of 3.2 mg/mL) may offer advantages over a low dose of propofol (with plasma concentrations of 1.8 mg/mL) for brain protection during CPB[. Previous studies have shown that OPCAB is better than conventional CABG by decreasing the release of S100β protein. Consequently, the pattern of S100β release at different stages of OPCAB procedures has become a valuable indicator of the early detection of neuronal clinical and subclinical injury[.

The present study revealed that CSF and serum S100 and S100β began to increase during CPB, peaked at the end of CPB for each indicator. However, CSF 100 showed a second peak at T7, and CSF S100β continued to be high until T4 and then gradually reduced. The results may indicate that S100 and S100β concentrations in the CSF are more sensitive than in the serum for indicating cerebral damage during cardiac surgery. CSF/serum S100 and S100β ratios may reflect the cerebral damage more accurately with a CSF-serum separation showing a sustained S100(β) release from the damaged brain tissues. The separation trends displayed from T5 for S100, and between T2 and T7 for S100β, respectively. This may hint that physiological and hemodynamic properties of the two proteins can be different and therefore showing distinct metabolic features after cardiac surgery. Intra-subgroup comparisons of serum S100(β) at T3 and T7 showed younger age, OPCAB, normothermia and positive intervene and even shorter CPB duration may reduce significantly the release of S100 and S100β. Serum ∆S100 and ∆S100β may also illustrate the severity of the cerebral damage during the operation. ∆S100, the difference between peak S100 and baseline S100, was reported to be 0.88 (0.48-3.23) in overall, 0.29 (0.18-0.44) in neonates and 1.1 (0.48-3.23) in infants[. In line with the results of serum S100(β) at T3 and T7, the study showed discrepancy of ∆S100 between aorta and congenital heart defect operations as well as extensive discrepancies of ∆S100β within age, operation, core temperature and cerebral damage subgroups. Despite the possible influence by the blood recovery transfusion, the indicators may still reflect the cerebral damage during cardiac surgery. In general, the release of S100 and S100β may correlate with age, operative method, CPB duration, core temperature and the application of intervenes during the operation. CSF S100(β) may be more reliable than serum S100(β), however, too aggressive drainage of CSF carries the risk of cerebral hernia and subdural hemorrhage[.

CONCLUSION

S100 and S100β in CSF can be more accurate than in the serum for the evaluations of cerebral damage in cardiac surgery. However, CSF biopsies are limited. But serum S100β and ∆S100β seems to be more sensitive than serum S100 and ∆S100. The cerebral damage in cardiac surgery might be associated with younger age, lower core temperature and longer CPB duration during the operation. Effective intervenes with modified CPB circuit filters or oxygenators and supplemented anesthetic agents or priming components may alleviate the cerebral damage.

Abbreviations, acronyms & symbols
CABG	Coronary artery bypass grafting
CPB	Cardiopulmonary bypass
CSF	Cerebrospinal fluid
OPCAB	Off-pump coronary artery bypass
POD	Postoperative day

Authors’ roles & responsibilities
SMY	Main Author